A priority control system for OR-parallel Prolog and its performance evaluation
In this paper we propose a method for performing priority control in OR-parallel execution of Prolog, with an evaluation function specified by the user as the priority. In the existing OR-parallel Prolog system there was a possibility that the number of goals executed in parallel could increase in combination, but by assigning priorities to the goals and executing in order of priority, the execution can be done in such a way that the search space does not have to be unnecessarily expanded during execution and only goals near the optimum solution are selected. The effectiveness of this method is demonstrated by implementing the processing system on a shared memory multiprocessor machine and showing that so-called super-linear speedup was obtained by applying to molecular phylogenetic analysis.
Unable to display preview. Download preview PDF.
- 1.Conery, J. S.: Parallel execution of logic programs. Kluwer Academic Publishers, Norwell, MA (1987) 35–61Google Scholar
- 3.Lusk, E. et al.: The Aurora OR-parallel Prolog system. New Generation Computing 8 (1990) 243–271Google Scholar
- 4.Warren, D. H. D.: An abstract Prolog instruction set. SRI Tech. Note 309 (1985)Google Scholar
- 5.Gupta, G., Jayaraman, B.: On criteria for OR-parallel execution models of logic programs. Proc. of NACLP'90 (1990) 737–756Google Scholar
- 6.Ichiyoshi, N.: Parallel implementation schemes of logic programming languages. J. of Information Processing Society of Japan 32 (1991) 435–449 (in Japanese)Google Scholar
- 7.Felsenstein, J.: Evolutionary trees from DNA sequences: A maximum likelihood approach. J. of Molecular Evolution 17 (1981) 368–376Google Scholar
- 8.Olsen, G. J., Matsuda, H., Hagstrom, R., Overbeek, R.: fastDNAml: A tool for construction of phylogenetic trees of DNA sequences using maximum likelihood. Computer Applications in Biological Sciences (1993) (in press)Google Scholar
- 9.Matsuda, H., Olsen, G. J., Hagstrom, R., Overbeek, R., Kaneda, Y.: Implementation of a parallel processing system for inference of phylogenetic trees. Proc. of Pac. Rim Conf. on Commun., Comput. and Signal Processing (1993) 280–283Google Scholar